This Forbes article is based on a discussion with Don McInnes who lead the Conversational Design team at Wells Fargo. Don suggested that the development of a conversational agent requires a broader set of learning data than is possible within a single bank and recommended Wells Fargo utilize a commercial product instead of building its own.
This is actually a common problem in training supervised machine learning models. The more data that can be used to train the model and the broader that data is related to the area of expertise under study, the faster the model learns and the better the model performs. This is why Google in search and ThreatMetrix in authentication have a market advantage over many other competitors.
Apparently the Wells Fargo research team thinks they have enough data and decided to go it alone. The good news is that if they run into trouble, this is a mistakes that can be easily corrected.
“He led the Conversational Design team at Wells Fargo for about a year before concluding the bank would lag behind because it wouldn’t take full advantage of commercially available software.
“I was frustrated because I couldn’t seem to help people understand how much value there is in learning the hard lessons of Conversational Design by using existing products. Things in this space are moving so quickly that unless you are a software company, specialized in – and dedicated to – refining CAI capabilities, you’re going to fall even further behind with every passing day,” he said.”
Overview by Tim Sloane, VP, Payments Innovation at Mercator Advisory Group